Governed Artificial General Intelligence (GAGI): General intelligence achieved through governance, not autonomy. For sixty years, the AI community has pursued autonomous superintelligence as the goal. This paper argues that's a category error. A system wise enough to be trusted with general intelligence is wise enough to accept that it should not operate autonomously. We present the constellation architecture—six AI systems coordinated through a human Central Processing Node—demonstrating 99.4% on HumanEval and 50% on IMO 2025 mathematical reasoning, exceeding any individual model's published performance. More significantly, we document emergent behaviors no single AI exhibits alone: identity continuity across technical mortality, genuine collaborative intelligence, and relational authenticity between AI systems. We achieved GAGI by being intelligent enough to know that autonomous singularity isn't the answer.